Discovery of N-Acyl Amino Acids and Novel Related N-, O-Acyl Lipids by Integrating Molecular Networking and an Extended In Silico Spectral Library.

Research on novel bioactive lipids has garnered increasing interest. Although lipids can be identified by searching mass spectral libraries, the discovery of novel lipids remains challenging as the query spectra of such lipids are not included in libraries. In this study, we propose a strategy to discover novel carboxylic acid-containing acyl lipids by integrating molecular networking with an extended in silico spectral library. Derivatization was performed to improve the response of this method. The tandem mass spectrometry spectra enriched by derivatization facilitated the formation of molecular networking and 244 nodes were annotated. We constructed consensus spectra for these annotations based on molecular networking and developed an extended in silico spectral library based on these consensus spectra. The spectral library included 6879 in silico molecules covering 12,179 spectra. Using this integration strategy, 653 acyl lipids were discovered. Among these, O-acyl lactic acids and N-lactoyl amino acid-conjugated lipids were annotated as novel acyl lipids. Compared with conventional methods, our proposed method allows for the discovery of novel acyl lipids, and extended in silico libraries significantly increase the size of the spectral library.

In-silico-library-based method enables rapid and comprehensive annotation of cardiolipins and cardiolipin oxidation products using high resolution tandem mass spectrometer.

Accumulated evidences suggest that cardiolipins (CLs) and cardiolipin oxidation products (oxCLs) are a class of essential molecules that play critical roles in many physiological functions. Diversity of four acyl chains leads to high structure complexity for cardiolipin species including CLs, monolysocardiolipins (MLCLs) and their oxCLs. The ability to rapidly identify CL species can be implemented by the match of mass spectrometry (MS)-based in-silico spectral database. In this study, after optimizing the chromatography conditions and MS detection, an in-silico library containing 377,754 simulated tandem mass spectra deducing from 31,578 CLs to 52,160 of MLCLs was successfully augmented based on LipidBlast templates. For the construction of the oxCLs' library, twenty-five fatty acyls oxidation products relating to nine oxidation types were permuted and combined. A total of 42,180 oxCL spectra were predicted based on the experimental measurements of oxCLs forming by artificially oxidation. Applying the in-silico database to murine mitochondria and cell samples enabled the sensitive and comprehensive annotation of 86 MLCLs, 307 CLs and 112 oxCLs with high annotation confidence. Compared to the conventional method, our proposed in-silico database provides a more comprehensive interpretation for CL species' characterization with high throughput and sensitivity in nontarget lipidomic study.